In [20]:
sns.catplot(data = interim_sqrt_df.loc[interim_sqrt_df['sess']==select_sess], x='interim_pass', kind='count')
plt.show()
select_sess = 3
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import statsmodels.formula.api as smf
import statsmodels.api as sm
from sklearn.metrics import roc_curve
from sklearn.metrics import roc_auc_score
from patsy import dmatrices
from numpy import arange
from sklearn.decomposition import PCA
from sklearn.cluster import KMeans
from sklearn.metrics import silhouette_score
from scipy.cluster import hierarchy
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder
from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import PolynomialFeatures
from sklearn.linear_model import ElasticNet
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import StratifiedKFold
from sklearn.model_selection import GridSearchCV
import time
CMPINF2120_EPM_FUNC_INCL_Over_Lisa.ipynb includes functions used in this notebook.
%run CMPINF2120_EPM_FUNC_INCL_Over_Lisa.ipynb
interim_sqrt_path = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/tp_sqrt_inputs_interim_df.csv'
interim_sqrt_init = pd.read_csv(interim_sqrt_path)
interim_sqrt_init.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 3642 entries, 0 to 3641 Data columns (total 83 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 sess 3642 non-null int64 1 sid 3642 non-null int64 2 actv_grp 3642 non-null object 3 total_ms_tp000_sqrt 3642 non-null float64 4 mw_tp000_sqrt 3642 non-null float64 5 mwc_tp000_sqrt 3642 non-null float64 6 mcl_tp000_sqrt 3642 non-null float64 7 mcr_tp000_sqrt 3642 non-null float64 8 mm_tp000_sqrt 3642 non-null float64 9 ks_tp000_sqrt 3642 non-null float64 10 total_ms_tp010_sqrt 3642 non-null float64 11 mw_tp010_sqrt 3642 non-null float64 12 mwc_tp010_sqrt 3642 non-null float64 13 mcl_tp010_sqrt 3642 non-null float64 14 mcr_tp010_sqrt 3642 non-null float64 15 mm_tp010_sqrt 3642 non-null float64 16 ks_tp010_sqrt 3642 non-null float64 17 total_ms_tp020_sqrt 3642 non-null float64 18 mw_tp020_sqrt 3642 non-null float64 19 mwc_tp020_sqrt 3642 non-null float64 20 mcl_tp020_sqrt 3642 non-null float64 21 mcr_tp020_sqrt 3642 non-null float64 22 mm_tp020_sqrt 3642 non-null float64 23 ks_tp020_sqrt 3642 non-null float64 24 total_ms_tp030_sqrt 3642 non-null float64 25 mw_tp030_sqrt 3642 non-null float64 26 mwc_tp030_sqrt 3642 non-null float64 27 mcl_tp030_sqrt 3642 non-null float64 28 mcr_tp030_sqrt 3642 non-null float64 29 mm_tp030_sqrt 3642 non-null float64 30 ks_tp030_sqrt 3642 non-null float64 31 total_ms_tp040_sqrt 3642 non-null float64 32 mw_tp040_sqrt 3642 non-null float64 33 mwc_tp040_sqrt 3642 non-null float64 34 mcl_tp040_sqrt 3642 non-null float64 35 mcr_tp040_sqrt 3642 non-null float64 36 mm_tp040_sqrt 3642 non-null float64 37 ks_tp040_sqrt 3642 non-null float64 38 total_ms_tp050_sqrt 3642 non-null float64 39 mw_tp050_sqrt 3642 non-null float64 40 mwc_tp050_sqrt 3642 non-null float64 41 mcl_tp050_sqrt 3642 non-null float64 42 mcr_tp050_sqrt 3642 non-null float64 43 mm_tp050_sqrt 3642 non-null float64 44 ks_tp050_sqrt 3642 non-null float64 45 total_ms_tp060_sqrt 3642 non-null float64 46 mw_tp060_sqrt 3642 non-null float64 47 mwc_tp060_sqrt 3642 non-null float64 48 mcl_tp060_sqrt 3642 non-null float64 49 mcr_tp060_sqrt 3642 non-null float64 50 mm_tp060_sqrt 3642 non-null float64 51 ks_tp060_sqrt 3642 non-null float64 52 total_ms_tp070_sqrt 3642 non-null float64 53 mw_tp070_sqrt 3642 non-null float64 54 mwc_tp070_sqrt 3642 non-null float64 55 mcl_tp070_sqrt 3642 non-null float64 56 mcr_tp070_sqrt 3642 non-null float64 57 mm_tp070_sqrt 3642 non-null float64 58 ks_tp070_sqrt 3642 non-null float64 59 total_ms_tp080_sqrt 3642 non-null float64 60 mw_tp080_sqrt 3642 non-null float64 61 mwc_tp080_sqrt 3642 non-null float64 62 mcl_tp080_sqrt 3642 non-null float64 63 mcr_tp080_sqrt 3642 non-null float64 64 mm_tp080_sqrt 3642 non-null float64 65 ks_tp080_sqrt 3642 non-null float64 66 total_ms_tp090_sqrt 3642 non-null float64 67 mw_tp090_sqrt 3642 non-null float64 68 mwc_tp090_sqrt 3642 non-null float64 69 mcl_tp090_sqrt 3642 non-null float64 70 mcr_tp090_sqrt 3642 non-null float64 71 mm_tp090_sqrt 3642 non-null float64 72 ks_tp090_sqrt 3642 non-null float64 73 total_ms_tp100_sqrt 3642 non-null float64 74 mw_tp100_sqrt 3642 non-null float64 75 mwc_tp100_sqrt 3642 non-null float64 76 mcl_tp100_sqrt 3642 non-null float64 77 mcr_tp100_sqrt 3642 non-null float64 78 mm_tp100_sqrt 3642 non-null float64 79 ks_tp100_sqrt 3642 non-null float64 80 interim_scr 3642 non-null float64 81 max_interim_scr 3642 non-null float64 82 interim_pass 3642 non-null float64 dtypes: float64(80), int64(2), object(1) memory usage: 2.3+ MB
interim_sqrt_init.isna().sum()
sess 0
sid 0
actv_grp 0
total_ms_tp000_sqrt 0
mw_tp000_sqrt 0
..
mm_tp100_sqrt 0
ks_tp100_sqrt 0
interim_scr 0
max_interim_scr 0
interim_pass 0
Length: 83, dtype: int64
interim_sqrt_init['sid'] = interim_sqrt_init['sid'].astype('object') interim_sqrt_init['sess'] = interim_sqrt_init['sess'].astype('object')
interim_sqrt_df = interim_sqrt_init.copy()
sqrt_vars = get_var_list(interim_sqrt_df,['sqrt'])
totl_vars = get_var_list_b(interim_sqrt_df,['total'])
mw_vars = get_var_list_b(interim_sqrt_df,['mw_'])
mwc_vars = get_var_list_b(interim_sqrt_df,['mwc'])
mcl_vars = get_var_list_b(interim_sqrt_df,['mcl'])
mcr_vars = get_var_list_b(interim_sqrt_df,['mcr'])
mm_vars = get_var_list_b(interim_sqrt_df,['mm'])
ks_vars = get_var_list_b(interim_sqrt_df,['ks'])
features_df = interim_sqrt_df.loc[:, sqrt_vars].copy()
features_df.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 3642 entries, 0 to 3641 Data columns (total 77 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 total_ms_tp000_sqrt 3642 non-null float64 1 mw_tp000_sqrt 3642 non-null float64 2 mwc_tp000_sqrt 3642 non-null float64 3 mcl_tp000_sqrt 3642 non-null float64 4 mcr_tp000_sqrt 3642 non-null float64 5 mm_tp000_sqrt 3642 non-null float64 6 ks_tp000_sqrt 3642 non-null float64 7 total_ms_tp010_sqrt 3642 non-null float64 8 mw_tp010_sqrt 3642 non-null float64 9 mwc_tp010_sqrt 3642 non-null float64 10 mcl_tp010_sqrt 3642 non-null float64 11 mcr_tp010_sqrt 3642 non-null float64 12 mm_tp010_sqrt 3642 non-null float64 13 ks_tp010_sqrt 3642 non-null float64 14 total_ms_tp020_sqrt 3642 non-null float64 15 mw_tp020_sqrt 3642 non-null float64 16 mwc_tp020_sqrt 3642 non-null float64 17 mcl_tp020_sqrt 3642 non-null float64 18 mcr_tp020_sqrt 3642 non-null float64 19 mm_tp020_sqrt 3642 non-null float64 20 ks_tp020_sqrt 3642 non-null float64 21 total_ms_tp030_sqrt 3642 non-null float64 22 mw_tp030_sqrt 3642 non-null float64 23 mwc_tp030_sqrt 3642 non-null float64 24 mcl_tp030_sqrt 3642 non-null float64 25 mcr_tp030_sqrt 3642 non-null float64 26 mm_tp030_sqrt 3642 non-null float64 27 ks_tp030_sqrt 3642 non-null float64 28 total_ms_tp040_sqrt 3642 non-null float64 29 mw_tp040_sqrt 3642 non-null float64 30 mwc_tp040_sqrt 3642 non-null float64 31 mcl_tp040_sqrt 3642 non-null float64 32 mcr_tp040_sqrt 3642 non-null float64 33 mm_tp040_sqrt 3642 non-null float64 34 ks_tp040_sqrt 3642 non-null float64 35 total_ms_tp050_sqrt 3642 non-null float64 36 mw_tp050_sqrt 3642 non-null float64 37 mwc_tp050_sqrt 3642 non-null float64 38 mcl_tp050_sqrt 3642 non-null float64 39 mcr_tp050_sqrt 3642 non-null float64 40 mm_tp050_sqrt 3642 non-null float64 41 ks_tp050_sqrt 3642 non-null float64 42 total_ms_tp060_sqrt 3642 non-null float64 43 mw_tp060_sqrt 3642 non-null float64 44 mwc_tp060_sqrt 3642 non-null float64 45 mcl_tp060_sqrt 3642 non-null float64 46 mcr_tp060_sqrt 3642 non-null float64 47 mm_tp060_sqrt 3642 non-null float64 48 ks_tp060_sqrt 3642 non-null float64 49 total_ms_tp070_sqrt 3642 non-null float64 50 mw_tp070_sqrt 3642 non-null float64 51 mwc_tp070_sqrt 3642 non-null float64 52 mcl_tp070_sqrt 3642 non-null float64 53 mcr_tp070_sqrt 3642 non-null float64 54 mm_tp070_sqrt 3642 non-null float64 55 ks_tp070_sqrt 3642 non-null float64 56 total_ms_tp080_sqrt 3642 non-null float64 57 mw_tp080_sqrt 3642 non-null float64 58 mwc_tp080_sqrt 3642 non-null float64 59 mcl_tp080_sqrt 3642 non-null float64 60 mcr_tp080_sqrt 3642 non-null float64 61 mm_tp080_sqrt 3642 non-null float64 62 ks_tp080_sqrt 3642 non-null float64 63 total_ms_tp090_sqrt 3642 non-null float64 64 mw_tp090_sqrt 3642 non-null float64 65 mwc_tp090_sqrt 3642 non-null float64 66 mcl_tp090_sqrt 3642 non-null float64 67 mcr_tp090_sqrt 3642 non-null float64 68 mm_tp090_sqrt 3642 non-null float64 69 ks_tp090_sqrt 3642 non-null float64 70 total_ms_tp100_sqrt 3642 non-null float64 71 mw_tp100_sqrt 3642 non-null float64 72 mwc_tp100_sqrt 3642 non-null float64 73 mcl_tp100_sqrt 3642 non-null float64 74 mcr_tp100_sqrt 3642 non-null float64 75 mm_tp100_sqrt 3642 non-null float64 76 ks_tp100_sqrt 3642 non-null float64 dtypes: float64(77) memory usage: 2.1 MB
feature_names = features_df.columns
len(feature_names)
77
interim_sqrt_df.info()
<class 'pandas.core.frame.DataFrame'> RangeIndex: 3642 entries, 0 to 3641 Data columns (total 83 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 sess 3642 non-null int64 1 sid 3642 non-null int64 2 actv_grp 3642 non-null object 3 total_ms_tp000_sqrt 3642 non-null float64 4 mw_tp000_sqrt 3642 non-null float64 5 mwc_tp000_sqrt 3642 non-null float64 6 mcl_tp000_sqrt 3642 non-null float64 7 mcr_tp000_sqrt 3642 non-null float64 8 mm_tp000_sqrt 3642 non-null float64 9 ks_tp000_sqrt 3642 non-null float64 10 total_ms_tp010_sqrt 3642 non-null float64 11 mw_tp010_sqrt 3642 non-null float64 12 mwc_tp010_sqrt 3642 non-null float64 13 mcl_tp010_sqrt 3642 non-null float64 14 mcr_tp010_sqrt 3642 non-null float64 15 mm_tp010_sqrt 3642 non-null float64 16 ks_tp010_sqrt 3642 non-null float64 17 total_ms_tp020_sqrt 3642 non-null float64 18 mw_tp020_sqrt 3642 non-null float64 19 mwc_tp020_sqrt 3642 non-null float64 20 mcl_tp020_sqrt 3642 non-null float64 21 mcr_tp020_sqrt 3642 non-null float64 22 mm_tp020_sqrt 3642 non-null float64 23 ks_tp020_sqrt 3642 non-null float64 24 total_ms_tp030_sqrt 3642 non-null float64 25 mw_tp030_sqrt 3642 non-null float64 26 mwc_tp030_sqrt 3642 non-null float64 27 mcl_tp030_sqrt 3642 non-null float64 28 mcr_tp030_sqrt 3642 non-null float64 29 mm_tp030_sqrt 3642 non-null float64 30 ks_tp030_sqrt 3642 non-null float64 31 total_ms_tp040_sqrt 3642 non-null float64 32 mw_tp040_sqrt 3642 non-null float64 33 mwc_tp040_sqrt 3642 non-null float64 34 mcl_tp040_sqrt 3642 non-null float64 35 mcr_tp040_sqrt 3642 non-null float64 36 mm_tp040_sqrt 3642 non-null float64 37 ks_tp040_sqrt 3642 non-null float64 38 total_ms_tp050_sqrt 3642 non-null float64 39 mw_tp050_sqrt 3642 non-null float64 40 mwc_tp050_sqrt 3642 non-null float64 41 mcl_tp050_sqrt 3642 non-null float64 42 mcr_tp050_sqrt 3642 non-null float64 43 mm_tp050_sqrt 3642 non-null float64 44 ks_tp050_sqrt 3642 non-null float64 45 total_ms_tp060_sqrt 3642 non-null float64 46 mw_tp060_sqrt 3642 non-null float64 47 mwc_tp060_sqrt 3642 non-null float64 48 mcl_tp060_sqrt 3642 non-null float64 49 mcr_tp060_sqrt 3642 non-null float64 50 mm_tp060_sqrt 3642 non-null float64 51 ks_tp060_sqrt 3642 non-null float64 52 total_ms_tp070_sqrt 3642 non-null float64 53 mw_tp070_sqrt 3642 non-null float64 54 mwc_tp070_sqrt 3642 non-null float64 55 mcl_tp070_sqrt 3642 non-null float64 56 mcr_tp070_sqrt 3642 non-null float64 57 mm_tp070_sqrt 3642 non-null float64 58 ks_tp070_sqrt 3642 non-null float64 59 total_ms_tp080_sqrt 3642 non-null float64 60 mw_tp080_sqrt 3642 non-null float64 61 mwc_tp080_sqrt 3642 non-null float64 62 mcl_tp080_sqrt 3642 non-null float64 63 mcr_tp080_sqrt 3642 non-null float64 64 mm_tp080_sqrt 3642 non-null float64 65 ks_tp080_sqrt 3642 non-null float64 66 total_ms_tp090_sqrt 3642 non-null float64 67 mw_tp090_sqrt 3642 non-null float64 68 mwc_tp090_sqrt 3642 non-null float64 69 mcl_tp090_sqrt 3642 non-null float64 70 mcr_tp090_sqrt 3642 non-null float64 71 mm_tp090_sqrt 3642 non-null float64 72 ks_tp090_sqrt 3642 non-null float64 73 total_ms_tp100_sqrt 3642 non-null float64 74 mw_tp100_sqrt 3642 non-null float64 75 mwc_tp100_sqrt 3642 non-null float64 76 mcl_tp100_sqrt 3642 non-null float64 77 mcr_tp100_sqrt 3642 non-null float64 78 mm_tp100_sqrt 3642 non-null float64 79 ks_tp100_sqrt 3642 non-null float64 80 interim_scr 3642 non-null float64 81 max_interim_scr 3642 non-null float64 82 interim_pass 3642 non-null float64 dtypes: float64(80), int64(2), object(1) memory usage: 2.3+ MB
sns.catplot(data = interim_sqrt_df.loc[interim_sqrt_df['sess']==select_sess], x='interim_pass', kind='count')
plt.show()
select_sess_df = interim_sqrt_df.loc[interim_sqrt_df['sess']==select_sess].drop(columns=['interim_scr','max_interim_scr']).copy()
select_sess_df.info()
<class 'pandas.core.frame.DataFrame'> Int64Index: 684 entries, 653 to 1336 Data columns (total 81 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 sess 684 non-null int64 1 sid 684 non-null int64 2 actv_grp 684 non-null object 3 total_ms_tp000_sqrt 684 non-null float64 4 mw_tp000_sqrt 684 non-null float64 5 mwc_tp000_sqrt 684 non-null float64 6 mcl_tp000_sqrt 684 non-null float64 7 mcr_tp000_sqrt 684 non-null float64 8 mm_tp000_sqrt 684 non-null float64 9 ks_tp000_sqrt 684 non-null float64 10 total_ms_tp010_sqrt 684 non-null float64 11 mw_tp010_sqrt 684 non-null float64 12 mwc_tp010_sqrt 684 non-null float64 13 mcl_tp010_sqrt 684 non-null float64 14 mcr_tp010_sqrt 684 non-null float64 15 mm_tp010_sqrt 684 non-null float64 16 ks_tp010_sqrt 684 non-null float64 17 total_ms_tp020_sqrt 684 non-null float64 18 mw_tp020_sqrt 684 non-null float64 19 mwc_tp020_sqrt 684 non-null float64 20 mcl_tp020_sqrt 684 non-null float64 21 mcr_tp020_sqrt 684 non-null float64 22 mm_tp020_sqrt 684 non-null float64 23 ks_tp020_sqrt 684 non-null float64 24 total_ms_tp030_sqrt 684 non-null float64 25 mw_tp030_sqrt 684 non-null float64 26 mwc_tp030_sqrt 684 non-null float64 27 mcl_tp030_sqrt 684 non-null float64 28 mcr_tp030_sqrt 684 non-null float64 29 mm_tp030_sqrt 684 non-null float64 30 ks_tp030_sqrt 684 non-null float64 31 total_ms_tp040_sqrt 684 non-null float64 32 mw_tp040_sqrt 684 non-null float64 33 mwc_tp040_sqrt 684 non-null float64 34 mcl_tp040_sqrt 684 non-null float64 35 mcr_tp040_sqrt 684 non-null float64 36 mm_tp040_sqrt 684 non-null float64 37 ks_tp040_sqrt 684 non-null float64 38 total_ms_tp050_sqrt 684 non-null float64 39 mw_tp050_sqrt 684 non-null float64 40 mwc_tp050_sqrt 684 non-null float64 41 mcl_tp050_sqrt 684 non-null float64 42 mcr_tp050_sqrt 684 non-null float64 43 mm_tp050_sqrt 684 non-null float64 44 ks_tp050_sqrt 684 non-null float64 45 total_ms_tp060_sqrt 684 non-null float64 46 mw_tp060_sqrt 684 non-null float64 47 mwc_tp060_sqrt 684 non-null float64 48 mcl_tp060_sqrt 684 non-null float64 49 mcr_tp060_sqrt 684 non-null float64 50 mm_tp060_sqrt 684 non-null float64 51 ks_tp060_sqrt 684 non-null float64 52 total_ms_tp070_sqrt 684 non-null float64 53 mw_tp070_sqrt 684 non-null float64 54 mwc_tp070_sqrt 684 non-null float64 55 mcl_tp070_sqrt 684 non-null float64 56 mcr_tp070_sqrt 684 non-null float64 57 mm_tp070_sqrt 684 non-null float64 58 ks_tp070_sqrt 684 non-null float64 59 total_ms_tp080_sqrt 684 non-null float64 60 mw_tp080_sqrt 684 non-null float64 61 mwc_tp080_sqrt 684 non-null float64 62 mcl_tp080_sqrt 684 non-null float64 63 mcr_tp080_sqrt 684 non-null float64 64 mm_tp080_sqrt 684 non-null float64 65 ks_tp080_sqrt 684 non-null float64 66 total_ms_tp090_sqrt 684 non-null float64 67 mw_tp090_sqrt 684 non-null float64 68 mwc_tp090_sqrt 684 non-null float64 69 mcl_tp090_sqrt 684 non-null float64 70 mcr_tp090_sqrt 684 non-null float64 71 mm_tp090_sqrt 684 non-null float64 72 ks_tp090_sqrt 684 non-null float64 73 total_ms_tp100_sqrt 684 non-null float64 74 mw_tp100_sqrt 684 non-null float64 75 mwc_tp100_sqrt 684 non-null float64 76 mcl_tp100_sqrt 684 non-null float64 77 mcr_tp100_sqrt 684 non-null float64 78 mm_tp100_sqrt 684 non-null float64 79 ks_tp100_sqrt 684 non-null float64 80 interim_pass 684 non-null float64 dtypes: float64(78), int64(2), object(1) memory usage: 438.2+ KB
select_sess_df.isna().sum()
sess 0
sid 0
actv_grp 0
total_ms_tp000_sqrt 0
mw_tp000_sqrt 0
..
mcl_tp100_sqrt 0
mcr_tp100_sqrt 0
mm_tp100_sqrt 0
ks_tp100_sqrt 0
interim_pass 0
Length: 81, dtype: int64
features_df = select_sess_df.loc[:, sqrt_vars].copy()
features_df.info()
<class 'pandas.core.frame.DataFrame'> Int64Index: 684 entries, 653 to 1336 Data columns (total 77 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 total_ms_tp000_sqrt 684 non-null float64 1 mw_tp000_sqrt 684 non-null float64 2 mwc_tp000_sqrt 684 non-null float64 3 mcl_tp000_sqrt 684 non-null float64 4 mcr_tp000_sqrt 684 non-null float64 5 mm_tp000_sqrt 684 non-null float64 6 ks_tp000_sqrt 684 non-null float64 7 total_ms_tp010_sqrt 684 non-null float64 8 mw_tp010_sqrt 684 non-null float64 9 mwc_tp010_sqrt 684 non-null float64 10 mcl_tp010_sqrt 684 non-null float64 11 mcr_tp010_sqrt 684 non-null float64 12 mm_tp010_sqrt 684 non-null float64 13 ks_tp010_sqrt 684 non-null float64 14 total_ms_tp020_sqrt 684 non-null float64 15 mw_tp020_sqrt 684 non-null float64 16 mwc_tp020_sqrt 684 non-null float64 17 mcl_tp020_sqrt 684 non-null float64 18 mcr_tp020_sqrt 684 non-null float64 19 mm_tp020_sqrt 684 non-null float64 20 ks_tp020_sqrt 684 non-null float64 21 total_ms_tp030_sqrt 684 non-null float64 22 mw_tp030_sqrt 684 non-null float64 23 mwc_tp030_sqrt 684 non-null float64 24 mcl_tp030_sqrt 684 non-null float64 25 mcr_tp030_sqrt 684 non-null float64 26 mm_tp030_sqrt 684 non-null float64 27 ks_tp030_sqrt 684 non-null float64 28 total_ms_tp040_sqrt 684 non-null float64 29 mw_tp040_sqrt 684 non-null float64 30 mwc_tp040_sqrt 684 non-null float64 31 mcl_tp040_sqrt 684 non-null float64 32 mcr_tp040_sqrt 684 non-null float64 33 mm_tp040_sqrt 684 non-null float64 34 ks_tp040_sqrt 684 non-null float64 35 total_ms_tp050_sqrt 684 non-null float64 36 mw_tp050_sqrt 684 non-null float64 37 mwc_tp050_sqrt 684 non-null float64 38 mcl_tp050_sqrt 684 non-null float64 39 mcr_tp050_sqrt 684 non-null float64 40 mm_tp050_sqrt 684 non-null float64 41 ks_tp050_sqrt 684 non-null float64 42 total_ms_tp060_sqrt 684 non-null float64 43 mw_tp060_sqrt 684 non-null float64 44 mwc_tp060_sqrt 684 non-null float64 45 mcl_tp060_sqrt 684 non-null float64 46 mcr_tp060_sqrt 684 non-null float64 47 mm_tp060_sqrt 684 non-null float64 48 ks_tp060_sqrt 684 non-null float64 49 total_ms_tp070_sqrt 684 non-null float64 50 mw_tp070_sqrt 684 non-null float64 51 mwc_tp070_sqrt 684 non-null float64 52 mcl_tp070_sqrt 684 non-null float64 53 mcr_tp070_sqrt 684 non-null float64 54 mm_tp070_sqrt 684 non-null float64 55 ks_tp070_sqrt 684 non-null float64 56 total_ms_tp080_sqrt 684 non-null float64 57 mw_tp080_sqrt 684 non-null float64 58 mwc_tp080_sqrt 684 non-null float64 59 mcl_tp080_sqrt 684 non-null float64 60 mcr_tp080_sqrt 684 non-null float64 61 mm_tp080_sqrt 684 non-null float64 62 ks_tp080_sqrt 684 non-null float64 63 total_ms_tp090_sqrt 684 non-null float64 64 mw_tp090_sqrt 684 non-null float64 65 mwc_tp090_sqrt 684 non-null float64 66 mcl_tp090_sqrt 684 non-null float64 67 mcr_tp090_sqrt 684 non-null float64 68 mm_tp090_sqrt 684 non-null float64 69 ks_tp090_sqrt 684 non-null float64 70 total_ms_tp100_sqrt 684 non-null float64 71 mw_tp100_sqrt 684 non-null float64 72 mwc_tp100_sqrt 684 non-null float64 73 mcl_tp100_sqrt 684 non-null float64 74 mcr_tp100_sqrt 684 non-null float64 75 mm_tp100_sqrt 684 non-null float64 76 ks_tp100_sqrt 684 non-null float64 dtypes: float64(77) memory usage: 416.8 KB
feature_names = features_df.columns
len(feature_names)
77
sns.catplot(data = select_sess_df, kind='box', aspect=3.5)
plt.show()
fig, ax = plt.subplots(figsize=(12, 8))
sns.heatmap(data = select_sess_df[feature_names].corr(),
vmin=-1, vmax=1, center = 0,
cmap='coolwarm',
ax=ax)
plt.show()
%run CMPINF2120_EPM_PCA_INCL_Over_Lisa.ipynb
<class 'pandas.core.frame.DataFrame'> RangeIndex: 684 entries, 0 to 683 Data columns (total 77 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 PC01 684 non-null float64 1 PC02 684 non-null float64 2 PC03 684 non-null float64 3 PC04 684 non-null float64 4 PC05 684 non-null float64 5 PC06 684 non-null float64 6 PC07 684 non-null float64 7 PC08 684 non-null float64 8 PC09 684 non-null float64 9 PC10 684 non-null float64 10 PC11 684 non-null float64 11 PC12 684 non-null float64 12 PC13 684 non-null float64 13 PC14 684 non-null float64 14 PC15 684 non-null float64 15 PC16 684 non-null float64 16 PC17 684 non-null float64 17 PC18 684 non-null float64 18 PC19 684 non-null float64 19 PC20 684 non-null float64 20 PC21 684 non-null float64 21 PC22 684 non-null float64 22 PC23 684 non-null float64 23 PC24 684 non-null float64 24 PC25 684 non-null float64 25 PC26 684 non-null float64 26 PC27 684 non-null float64 27 PC28 684 non-null float64 28 PC29 684 non-null float64 29 PC30 684 non-null float64 30 PC31 684 non-null float64 31 PC32 684 non-null float64 32 PC33 684 non-null float64 33 PC34 684 non-null float64 34 PC35 684 non-null float64 35 PC36 684 non-null float64 36 PC37 684 non-null float64 37 PC38 684 non-null float64 38 PC39 684 non-null float64 39 PC40 684 non-null float64 40 PC41 684 non-null float64 41 PC42 684 non-null float64 42 PC43 684 non-null float64 43 PC44 684 non-null float64 44 PC45 684 non-null float64 45 PC46 684 non-null float64 46 PC47 684 non-null float64 47 PC48 684 non-null float64 48 PC49 684 non-null float64 49 PC50 684 non-null float64 50 PC51 684 non-null float64 51 PC52 684 non-null float64 52 PC53 684 non-null float64 53 PC54 684 non-null float64 54 PC55 684 non-null float64 55 PC56 684 non-null float64 56 PC57 684 non-null float64 57 PC58 684 non-null float64 58 PC59 684 non-null float64 59 PC60 684 non-null float64 60 PC61 684 non-null float64 61 PC62 684 non-null float64 62 PC63 684 non-null float64 63 PC64 684 non-null float64 64 PC65 684 non-null float64 65 PC66 684 non-null float64 66 PC67 684 non-null float64 67 PC68 684 non-null float64 68 PC69 684 non-null float64 69 PC70 684 non-null float64 70 PC71 684 non-null float64 71 PC72 684 non-null float64 72 PC73 684 non-null float64 73 PC74 684 non-null float64 74 PC75 684 non-null float64 75 PC76 684 non-null float64 76 PC77 684 non-null float64 dtypes: float64(77) memory usage: 411.6 KB
pc_scores_select_df = pc_scores_df.copy()
select_Xtimepoints = Xtimepoints.copy()
num_pcs = 12
pc_cluster_df = pd.concat([select_sess_df.loc[:,['sid','actv_grp','interim_pass']].copy(), pc_scores_select_df.copy()], axis=1)
pc_cluster_df.sid = pc_cluster_df.sid.astype('category')
sns.set_style('whitegrid')
sns.relplot(data = pc_cluster_df, x='PC01', y='PC02')
plt.show()
sns.set_style('whitegrid')
sns.relplot(data = pc_cluster_df, x='PC01', y='PC02', hue='actv_grp', legend=False)
plt.show()
sns.set_style('whitegrid')
sns.relplot(data = pc_cluster_df, x='PC01', y='PC02', hue='sid', legend=False)
plt.show()
sns.set_style('whitegrid')
sns.relplot(data = pc_cluster_df, x='PC01', y='PC02', hue='interim_pass', legend=False)
plt.show()
The optimal number of clusters is 4.
tots_within = []
K = range(1, 31)
for k in K:
km = KMeans(n_clusters=k, random_state=121, n_init=25, max_iter=500).fit( pc_scores_select_df )
tots_within.append( km.inertia_ )
pc_scores_12_df = pc_scores_select_df.iloc[:,:12].copy()
sns.catplot(data = pc_scores_12_df, kind='box')
<seaborn.axisgrid.FacetGrid at 0x7fbccccc2a60>
The silhouette coefficient shows that 2 clusters are optimal.
sil_coef = []
K = range(2, 31)
for k in K:
k_label = KMeans(n_clusters=k, random_state=121, n_init=25, max_iter=500).fit_predict( pc_scores_12_df )
sil_coef.append( silhouette_score(pc_scores_12_df, k_label) )
fig, ax = plt.subplots(figsize=(12, 8))
ax.plot(K, sil_coef, 'o-')
ax.set_xlabel('number of clusters')
ax.set_ylabel('average silhouette coefficient')
plt.show()
The optimal number of clusters chosen from the dendrogram is 3 or 8.
hclust_ward = hierarchy.ward( pc_scores_12_df )
fig = plt.figure( figsize=(18,6) )
dn = hierarchy.dendrogram( hclust_ward, no_labels=True )
plt.show()
ward_cut_a = hierarchy.cut_tree( hclust_ward, n_clusters=8 )
ward_cut_a0 = hierarchy.cut_tree( hclust_ward, n_clusters=3 )
ward_cut_a.shape
(684, 1)
type( ward_cut_a )
numpy.ndarray
ward_cut_a.ravel()
array([0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 2, 2, 3, 2, 2,
2, 2, 4, 4, 4, 4, 4, 4, 4, 4, 2, 2, 2, 3, 2, 2, 2, 2, 2, 2, 2, 3,
2, 3, 2, 2, 3, 3, 2, 3, 1, 2, 2, 2, 2, 2, 2, 3, 2, 2, 2, 2, 0, 5,
5, 0, 5, 0, 5, 5, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 0, 1, 1, 1, 1,
4, 4, 4, 4, 4, 4, 4, 4, 2, 2, 3, 3, 2, 2, 2, 2, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 1, 1, 5, 5, 5, 5, 5,
5, 5, 5, 2, 2, 2, 3, 2, 3, 1, 2, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0,
0, 0, 0, 0, 0, 3, 2, 2, 3, 2, 2, 2, 2, 1, 1, 0, 0, 1, 0, 1, 0, 1,
1, 1, 0, 1, 1, 1, 0, 2, 2, 2, 3, 2, 3, 2, 2, 2, 2, 2, 3, 3, 1, 2,
2, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 3, 2, 3, 2, 2, 0, 0, 0, 0, 0, 0,
0, 0, 5, 5, 5, 5, 5, 5, 5, 5, 3, 2, 3, 3, 1, 3, 3, 2, 2, 2, 2, 2,
2, 2, 2, 2, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 1,
1, 0, 0, 1, 0, 1, 1, 5, 5, 5, 5, 5, 5, 5, 5, 4, 2, 4, 4, 2, 4, 4,
4, 3, 2, 2, 3, 2, 3, 2, 2, 1, 1, 1, 0, 1, 1, 1, 0, 5, 5, 5, 5, 5,
5, 1, 5, 1, 0, 0, 0, 1, 0, 1, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
3, 2, 3, 2, 2, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 2,
2, 2, 2, 5, 2, 5, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1,
0, 2, 2, 2, 1, 2, 2, 1, 2, 5, 5, 5, 5, 5, 5, 5, 5, 2, 2, 2, 2, 2,
2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 6, 5, 5, 5, 1, 5, 5, 5, 1, 1, 1,
1, 1, 1, 1, 1, 2, 2, 3, 2, 2, 2, 1, 5, 5, 5, 1, 5, 1, 6, 5, 0, 0,
0, 0, 0, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4, 4, 4, 0, 0, 0, 0, 0, 0, 0,
0, 3, 2, 2, 3, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 7, 7, 7, 7, 7,
2, 7, 7, 5, 5, 5, 5, 5, 5, 5, 5, 3, 2, 2, 3, 2, 3, 2, 2, 6, 1, 0,
0, 1, 0, 1, 1, 2, 2, 2, 3, 2, 3, 2, 3, 2, 5, 5, 5, 5, 5, 5, 5, 5,
0, 1, 1, 0, 1, 1, 0, 1, 2, 2, 2, 3, 2, 2, 2, 2, 6, 1, 1, 1, 1, 1,
1, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 0, 1, 0, 0, 1,
0, 1, 0, 4, 4, 4, 4, 4, 4, 4, 4, 3, 2, 2, 3, 2, 2, 2, 2, 2, 2, 3,
3, 2, 2, 2, 2, 2, 2, 2, 3, 2, 2, 2, 2, 4, 4, 4, 4, 4, 4, 4, 3, 2,
2, 3, 2, 2, 2, 2, 2, 2, 2, 2, 1, 2, 1, 2, 2, 4, 4, 3, 4, 3, 2, 4,
5, 2, 2, 3, 2, 3, 2, 2, 1, 1, 1, 1, 1, 1, 1, 7, 7, 7, 7, 7, 7, 7,
7, 7])
pc_scores_12_df.index
RangeIndex(start=0, stop=684, step=1)
pc_scores_12_df['hclust_a'] = pd.Series( ward_cut_a.ravel(), index=pc_scores_12_df.index )
pc_scores_12_df['hclust_a'] = pc_scores_12_df.hclust_a.astype('category')
pc_scores_12_df['hclust_a0'] = pd.Series( ward_cut_a0.ravel(), index=pc_scores_12_df.index )
pc_scores_12_df['hclust_a0'] = pc_scores_12_df.hclust_a0.astype('category')
sns.catplot(data = pc_scores_12_df, x='hclust_a0', hue='hclust_a', kind='count')
<seaborn.axisgrid.FacetGrid at 0x7fbca85c6910>
sns.relplot(data = pc_scores_12_df, x='PC01', y='PC02', hue='hclust_a')
plt.show()
sns.relplot(data = pc_scores_12_df, x='PC01', y='PC02', col='hclust_a0', hue='hclust_a')
plt.show()
select_sess_df to pc_cluster_df dataset (merge).¶pc_scores_outp_df = pd.concat([select_sess_df.loc[:,['sid','actv_grp','interim_pass']].copy(), pc_scores_12_df.copy()], axis=1)
pc_scores_outp_df.head()
| sid | actv_grp | interim_pass | PC01 | PC02 | PC03 | PC04 | PC05 | PC06 | PC07 | PC08 | PC09 | PC10 | PC11 | PC12 | hclust_a | hclust_a0 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 653 | 2.0 | Aulaweb | 1.0 | 1.243888 | -0.194865 | -1.422462 | 1.754434 | -0.746909 | -1.231178 | 3.020375 | 1.364173 | 1.212701 | 0.779248 | 1.449629 | -0.206498 | 4 | 0 |
| 654 | 2.0 | Blank | 1.0 | 0.106988 | -0.740887 | -1.407201 | 1.263576 | -0.712784 | -1.212260 | 3.074962 | 0.591506 | 1.617775 | 0.889883 | 1.513398 | -0.458913 | 4 | 0 |
| 655 | 2.0 | Deeds | 1.0 | -2.708881 | -2.133565 | 4.170313 | 0.857947 | -2.167076 | -1.038828 | 1.471239 | -2.133631 | 2.420663 | 0.183425 | 0.299691 | -0.262321 | 3 | 0 |
| 656 | 2.0 | Diagram | 1.0 | -0.897608 | -1.107260 | -0.248005 | 1.569805 | -0.992424 | -1.095644 | 3.205990 | 2.263915 | 1.546776 | -0.788815 | 0.130550 | 0.315816 | 4 | 0 |
| 657 | 2.0 | Other | 1.0 | -6.222978 | -3.908513 | 4.465966 | 0.245749 | -1.607440 | -0.835841 | 2.693033 | -0.105511 | 2.138642 | 0.655484 | 0.007186 | -0.300847 | 3 | 0 |
pc_scores_outp_df.interim_pass.mean()
0.5833333333333334
pc_scores_outp_df.hclust_a.value_counts().reset_index()
| index | hclust_a | |
|---|---|---|
| 0 | 2 | 183 |
| 1 | 0 | 142 |
| 2 | 1 | 123 |
| 3 | 5 | 90 |
| 4 | 4 | 75 |
| 5 | 3 | 51 |
| 6 | 7 | 16 |
| 7 | 6 | 4 |
pc_scores_outp_df.hclust_a0.value_counts().reset_index()
| index | hclust_a0 | |
|---|---|---|
| 0 | 0 | 451 |
| 1 | 1 | 127 |
| 2 | 2 | 106 |
fig, ax = plt.subplots(figsize=(18,14))
sns.heatmap(pd.crosstab(pc_scores_outp_df.hclust_a, pc_scores_outp_df.interim_pass, margins=True),
annot=True, annot_kws={'size': 10}, fmt='g')
plt.show()
fig, ax = plt.subplots(figsize=(18,14))
sns.heatmap(pd.crosstab(pc_scores_outp_df.hclust_a0, pc_scores_outp_df.interim_pass, margins=True),
annot=True, annot_kws={'size': 10}, fmt='g')
plt.show()
pc_to_model_df = pd.concat([select_sess_df.loc[:,['sid','actv_grp','interim_pass']].copy(), pc_scores_select_df.copy()], axis=1)
pc_features = ['PC01','PC02','PC03','PC04','PC05','PC06','PC07','PC08','PC09','PC10','PC11','PC12']
pc_features_str = ''
for ix, x in enumerate(pc_features):
if ix == len(pc_features) - 1:
pc_features_str = pc_features_str + x
else:
pc_features_str = pc_features_str + x + ' + '
pc_features_str
'PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12'
pc_desc_formulas = ['interim_pass ~ sid'
,'interim_pass ~ sid + actv_grp'
,'interim_pass ~ sid + actv_grp + ' + pc_features_str
,'interim_pass ~ sid * (' + pc_features_str + ')'
,'interim_pass ~ sid * (actv_grp + ' + pc_features_str + ')'
]
pc_pred_formulas = ['interim_pass ~ ' + pc_features_str
,'interim_pass ~ (' + pc_features_str + ')**2'
,'interim_pass ~ actv_grp + ' + pc_features_str
,'interim_pass ~ actv_grp * (' + pc_features_str + ')'
,'interim_pass ~ actv_grp + (' + pc_features_str + ')**2'
,'interim_pass ~ actv_grp * (' + pc_features_str + ')**2'
]
formula_list = pc_desc_formulas + pc_pred_formulas
formula_list
['interim_pass ~ sid', 'interim_pass ~ sid + actv_grp', 'interim_pass ~ sid + actv_grp + PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12', 'interim_pass ~ sid * (PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12)', 'interim_pass ~ sid * (actv_grp + PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12)', 'interim_pass ~ PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12', 'interim_pass ~ (PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12)**2', 'interim_pass ~ actv_grp + PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12', 'interim_pass ~ actv_grp * (PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12)', 'interim_pass ~ actv_grp + (PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12)**2', 'interim_pass ~ actv_grp * (PC01 + PC02 + PC03 + PC04 + PC05 + PC06 + PC07 + PC08 + PC09 + PC10 + PC11 + PC12)**2']
sk_list = make_dmat(pc_to_model_df, formula_list)
model_dim = make_dim_df(pc_to_model_df, sk_list, formula_list)
model_dim
| model name | dimensions | number of obs | dim < obs | |
|---|---|---|---|---|
| 0 | 0 | 2 | 1337 | Yes |
| 1 | 1 | 11 | 1337 | Yes |
| 2 | 2 | 23 | 1337 | Yes |
| 3 | 3 | 26 | 1337 | Yes |
| 4 | 4 | 44 | 1337 | Yes |
| 5 | 5 | 13 | 1337 | Yes |
| 6 | 6 | 79 | 1337 | Yes |
| 7 | 7 | 22 | 1337 | Yes |
| 8 | 8 | 130 | 1337 | Yes |
| 9 | 9 | 88 | 1337 | Yes |
| 10 | 10 | 790 | 1337 | Yes |
def fit_model(dm_y, dm_X):
return sm.GLM(dm_y, dm_X, family=sm.families.Binomial()).fit(bic_llf=True)
pc_dm00_y, pc_dm00_X = dmatrices(formula_list[0], data=pc_to_model_df, return_type='dataframe')
pc_dm01_y, pc_dm01_X = dmatrices(formula_list[1], data=pc_to_model_df, return_type='dataframe')
pc_dm02_y, pc_dm02_X = dmatrices(formula_list[2], data=pc_to_model_df, return_type='dataframe')
pc_dm03_y, pc_dm03_X = dmatrices(formula_list[3], data=pc_to_model_df, return_type='dataframe')
pc_dm04_y, pc_dm04_X = dmatrices(formula_list[4], data=pc_to_model_df, return_type='dataframe')
pc_dm05_y, pc_dm05_X = dmatrices(formula_list[5], data=pc_to_model_df, return_type='dataframe')
pc_dm06_y, pc_dm06_X = dmatrices(formula_list[6], data=pc_to_model_df, return_type='dataframe')
pc_dm07_y, pc_dm07_X = dmatrices(formula_list[7], data=pc_to_model_df, return_type='dataframe')
pc_dm08_y, pc_dm08_X = dmatrices(formula_list[8], data=pc_to_model_df, return_type='dataframe')
pc_dm09_y, pc_dm09_X = dmatrices(formula_list[9], data=pc_to_model_df, return_type='dataframe')
pc_dm10_y, pc_dm10_X = dmatrices(formula_list[10], data=pc_to_model_df, return_type='dataframe')
model_list = []
model_list.append( fit_model(pc_dm00_y, pc_dm00_X) )
time.sleep(5)
model_list.append( fit_model(pc_dm01_y, pc_dm01_X) )
The following models have perfect separation.
#time.sleep(5)
#model_list.append( fit_model(pc_dm02_y, pc_dm02_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm03_y, pc_dm03_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm04_y, pc_dm04_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm05_y, pc_dm05_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm06_y, pc_dm06_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm07_y, pc_dm07_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm08_y, pc_dm08_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm09_y, pc_dm09_X) )
#time.sleep(5)
#model_list.append( fit_model(pc_dm10_y, pc_dm10_X) )
model_results = pd.DataFrame({'model_name': ['mod00','mod01'],
'AIC': [mod.aic for mod in model_list],
'BIC': [mod.bic for mod in model_list]})
/Users/lisaover/opt/anaconda3/envs/cmpinf2120/lib/python3.8/site-packages/statsmodels/genmod/generalized_linear_model.py:1799: FutureWarning: The bic value is computed using the deviance formula. After 0.13 this will change to the log-likelihood based formula. This change has no impact on the relative rank of models compared using BIC. You can directly access the log-likelihood version using the `bic_llf` attribute. You can suppress this message by calling statsmodels.genmod.generalized_linear_model.SET_USE_BIC_LLF with True to get the LLF-based version now or False to retainthe deviance version. warnings.warn(
model_results
| model_name | AIC | BIC | |
|---|---|---|---|
| 0 | mod00 | 926.552350 | -3529.514949 |
| 1 | mod01 | 941.658395 | -3473.657283 |
sns.relplot(data = model_results.melt(id_vars=['model_name']),
x='model_name',
y='value',
col='variable',
col_wrap=2,
facet_kws = {'sharey': False},
height=5, aspect=2)
plt.show()
select_sess_df.info()
<class 'pandas.core.frame.DataFrame'> Int64Index: 684 entries, 653 to 1336 Data columns (total 81 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 sess 684 non-null int64 1 sid 684 non-null int64 2 actv_grp 684 non-null object 3 total_ms_tp000_sqrt 684 non-null float64 4 mw_tp000_sqrt 684 non-null float64 5 mwc_tp000_sqrt 684 non-null float64 6 mcl_tp000_sqrt 684 non-null float64 7 mcr_tp000_sqrt 684 non-null float64 8 mm_tp000_sqrt 684 non-null float64 9 ks_tp000_sqrt 684 non-null float64 10 total_ms_tp010_sqrt 684 non-null float64 11 mw_tp010_sqrt 684 non-null float64 12 mwc_tp010_sqrt 684 non-null float64 13 mcl_tp010_sqrt 684 non-null float64 14 mcr_tp010_sqrt 684 non-null float64 15 mm_tp010_sqrt 684 non-null float64 16 ks_tp010_sqrt 684 non-null float64 17 total_ms_tp020_sqrt 684 non-null float64 18 mw_tp020_sqrt 684 non-null float64 19 mwc_tp020_sqrt 684 non-null float64 20 mcl_tp020_sqrt 684 non-null float64 21 mcr_tp020_sqrt 684 non-null float64 22 mm_tp020_sqrt 684 non-null float64 23 ks_tp020_sqrt 684 non-null float64 24 total_ms_tp030_sqrt 684 non-null float64 25 mw_tp030_sqrt 684 non-null float64 26 mwc_tp030_sqrt 684 non-null float64 27 mcl_tp030_sqrt 684 non-null float64 28 mcr_tp030_sqrt 684 non-null float64 29 mm_tp030_sqrt 684 non-null float64 30 ks_tp030_sqrt 684 non-null float64 31 total_ms_tp040_sqrt 684 non-null float64 32 mw_tp040_sqrt 684 non-null float64 33 mwc_tp040_sqrt 684 non-null float64 34 mcl_tp040_sqrt 684 non-null float64 35 mcr_tp040_sqrt 684 non-null float64 36 mm_tp040_sqrt 684 non-null float64 37 ks_tp040_sqrt 684 non-null float64 38 total_ms_tp050_sqrt 684 non-null float64 39 mw_tp050_sqrt 684 non-null float64 40 mwc_tp050_sqrt 684 non-null float64 41 mcl_tp050_sqrt 684 non-null float64 42 mcr_tp050_sqrt 684 non-null float64 43 mm_tp050_sqrt 684 non-null float64 44 ks_tp050_sqrt 684 non-null float64 45 total_ms_tp060_sqrt 684 non-null float64 46 mw_tp060_sqrt 684 non-null float64 47 mwc_tp060_sqrt 684 non-null float64 48 mcl_tp060_sqrt 684 non-null float64 49 mcr_tp060_sqrt 684 non-null float64 50 mm_tp060_sqrt 684 non-null float64 51 ks_tp060_sqrt 684 non-null float64 52 total_ms_tp070_sqrt 684 non-null float64 53 mw_tp070_sqrt 684 non-null float64 54 mwc_tp070_sqrt 684 non-null float64 55 mcl_tp070_sqrt 684 non-null float64 56 mcr_tp070_sqrt 684 non-null float64 57 mm_tp070_sqrt 684 non-null float64 58 ks_tp070_sqrt 684 non-null float64 59 total_ms_tp080_sqrt 684 non-null float64 60 mw_tp080_sqrt 684 non-null float64 61 mwc_tp080_sqrt 684 non-null float64 62 mcl_tp080_sqrt 684 non-null float64 63 mcr_tp080_sqrt 684 non-null float64 64 mm_tp080_sqrt 684 non-null float64 65 ks_tp080_sqrt 684 non-null float64 66 total_ms_tp090_sqrt 684 non-null float64 67 mw_tp090_sqrt 684 non-null float64 68 mwc_tp090_sqrt 684 non-null float64 69 mcl_tp090_sqrt 684 non-null float64 70 mcr_tp090_sqrt 684 non-null float64 71 mm_tp090_sqrt 684 non-null float64 72 ks_tp090_sqrt 684 non-null float64 73 total_ms_tp100_sqrt 684 non-null float64 74 mw_tp100_sqrt 684 non-null float64 75 mwc_tp100_sqrt 684 non-null float64 76 mcl_tp100_sqrt 684 non-null float64 77 mcr_tp100_sqrt 684 non-null float64 78 mm_tp100_sqrt 684 non-null float64 79 ks_tp100_sqrt 684 non-null float64 80 interim_pass 684 non-null float64 dtypes: float64(78), int64(2), object(1) memory usage: 454.4+ KB
categorical_inputs = ['sid','actv_grp']
sqrt_vars
['total_ms_tp000_sqrt', 'mw_tp000_sqrt', 'mwc_tp000_sqrt', 'mcl_tp000_sqrt', 'mcr_tp000_sqrt', 'mm_tp000_sqrt', 'ks_tp000_sqrt', 'total_ms_tp010_sqrt', 'mw_tp010_sqrt', 'mwc_tp010_sqrt', 'mcl_tp010_sqrt', 'mcr_tp010_sqrt', 'mm_tp010_sqrt', 'ks_tp010_sqrt', 'total_ms_tp020_sqrt', 'mw_tp020_sqrt', 'mwc_tp020_sqrt', 'mcl_tp020_sqrt', 'mcr_tp020_sqrt', 'mm_tp020_sqrt', 'ks_tp020_sqrt', 'total_ms_tp030_sqrt', 'mw_tp030_sqrt', 'mwc_tp030_sqrt', 'mcl_tp030_sqrt', 'mcr_tp030_sqrt', 'mm_tp030_sqrt', 'ks_tp030_sqrt', 'total_ms_tp040_sqrt', 'mw_tp040_sqrt', 'mwc_tp040_sqrt', 'mcl_tp040_sqrt', 'mcr_tp040_sqrt', 'mm_tp040_sqrt', 'ks_tp040_sqrt', 'total_ms_tp050_sqrt', 'mw_tp050_sqrt', 'mwc_tp050_sqrt', 'mcl_tp050_sqrt', 'mcr_tp050_sqrt', 'mm_tp050_sqrt', 'ks_tp050_sqrt', 'total_ms_tp060_sqrt', 'mw_tp060_sqrt', 'mwc_tp060_sqrt', 'mcl_tp060_sqrt', 'mcr_tp060_sqrt', 'mm_tp060_sqrt', 'ks_tp060_sqrt', 'total_ms_tp070_sqrt', 'mw_tp070_sqrt', 'mwc_tp070_sqrt', 'mcl_tp070_sqrt', 'mcr_tp070_sqrt', 'mm_tp070_sqrt', 'ks_tp070_sqrt', 'total_ms_tp080_sqrt', 'mw_tp080_sqrt', 'mwc_tp080_sqrt', 'mcl_tp080_sqrt', 'mcr_tp080_sqrt', 'mm_tp080_sqrt', 'ks_tp080_sqrt', 'total_ms_tp090_sqrt', 'mw_tp090_sqrt', 'mwc_tp090_sqrt', 'mcl_tp090_sqrt', 'mcr_tp090_sqrt', 'mm_tp090_sqrt', 'ks_tp090_sqrt', 'total_ms_tp100_sqrt', 'mw_tp100_sqrt', 'mwc_tp100_sqrt', 'mcl_tp100_sqrt', 'mcr_tp100_sqrt', 'mm_tp100_sqrt', 'ks_tp100_sqrt']
categorical_transform = Pipeline( steps = [ ('dummy', OneHotEncoder(drop='first')) ] )
numeric_transform_pca = Pipeline( steps=[('std_input', StandardScaler()),
('pca', PCA(n_components=num_pcs)),
('std_features', StandardScaler())])
complex_transform = ColumnTransformer( transformers=[('num', numeric_transform_pca, sqrt_vars),
('cat', categorical_transform, categorical_inputs)])
complex_transform.transformers
[('num',
Pipeline(steps=[('std_input', StandardScaler()), ('pca', PCA(n_components=12)),
('std_features', StandardScaler())]),
['total_ms_tp000_sqrt',
'mw_tp000_sqrt',
'mwc_tp000_sqrt',
'mcl_tp000_sqrt',
'mcr_tp000_sqrt',
'mm_tp000_sqrt',
'ks_tp000_sqrt',
'total_ms_tp010_sqrt',
'mw_tp010_sqrt',
'mwc_tp010_sqrt',
'mcl_tp010_sqrt',
'mcr_tp010_sqrt',
'mm_tp010_sqrt',
'ks_tp010_sqrt',
'total_ms_tp020_sqrt',
'mw_tp020_sqrt',
'mwc_tp020_sqrt',
'mcl_tp020_sqrt',
'mcr_tp020_sqrt',
'mm_tp020_sqrt',
'ks_tp020_sqrt',
'total_ms_tp030_sqrt',
'mw_tp030_sqrt',
'mwc_tp030_sqrt',
'mcl_tp030_sqrt',
'mcr_tp030_sqrt',
'mm_tp030_sqrt',
'ks_tp030_sqrt',
'total_ms_tp040_sqrt',
'mw_tp040_sqrt',
'mwc_tp040_sqrt',
'mcl_tp040_sqrt',
'mcr_tp040_sqrt',
'mm_tp040_sqrt',
'ks_tp040_sqrt',
'total_ms_tp050_sqrt',
'mw_tp050_sqrt',
'mwc_tp050_sqrt',
'mcl_tp050_sqrt',
'mcr_tp050_sqrt',
'mm_tp050_sqrt',
'ks_tp050_sqrt',
'total_ms_tp060_sqrt',
'mw_tp060_sqrt',
'mwc_tp060_sqrt',
'mcl_tp060_sqrt',
'mcr_tp060_sqrt',
'mm_tp060_sqrt',
'ks_tp060_sqrt',
'total_ms_tp070_sqrt',
'mw_tp070_sqrt',
'mwc_tp070_sqrt',
'mcl_tp070_sqrt',
'mcr_tp070_sqrt',
'mm_tp070_sqrt',
'ks_tp070_sqrt',
'total_ms_tp080_sqrt',
'mw_tp080_sqrt',
'mwc_tp080_sqrt',
'mcl_tp080_sqrt',
'mcr_tp080_sqrt',
'mm_tp080_sqrt',
'ks_tp080_sqrt',
'total_ms_tp090_sqrt',
'mw_tp090_sqrt',
'mwc_tp090_sqrt',
'mcl_tp090_sqrt',
'mcr_tp090_sqrt',
'mm_tp090_sqrt',
'ks_tp090_sqrt',
'total_ms_tp100_sqrt',
'mw_tp100_sqrt',
'mwc_tp100_sqrt',
'mcl_tp100_sqrt',
'mcr_tp100_sqrt',
'mm_tp100_sqrt',
'ks_tp100_sqrt']),
('cat',
Pipeline(steps=[('dummy', OneHotEncoder(drop='first'))]),
['sid', 'actv_grp'])]
additive_transform = Pipeline( steps=[ ('prepro', complex_transform),
('make_poly', PolynomialFeatures(degree=1, interaction_only=True, include_bias=False))])
additive_transform.fit_transform(select_sess_df).shape
(684, 106)
len(sqrt_vars)
77
select_sess_df.shape
(684, 81)
select_sess_X = select_sess_df.drop(columns=['sess','interim_pass']).copy()
select_sess_df.interim_pass.value_counts()
1.0 399 0.0 285 Name: interim_pass, dtype: int64
select_sess_df.interim_pass.to_numpy().ravel()
array([1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 1.,
1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1.,
1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0.,
0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0.,
0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1.,
1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 1., 1., 1., 1., 1.,
1., 1., 1., 1.])
select_sess_y = select_sess_df.interim_pass.to_numpy().ravel()
select_sess_X.shape
(684, 79)
select_sess_X.columns
Index(['sid', 'actv_grp', 'total_ms_tp000_sqrt', 'mw_tp000_sqrt',
'mwc_tp000_sqrt', 'mcl_tp000_sqrt', 'mcr_tp000_sqrt', 'mm_tp000_sqrt',
'ks_tp000_sqrt', 'total_ms_tp010_sqrt', 'mw_tp010_sqrt',
'mwc_tp010_sqrt', 'mcl_tp010_sqrt', 'mcr_tp010_sqrt', 'mm_tp010_sqrt',
'ks_tp010_sqrt', 'total_ms_tp020_sqrt', 'mw_tp020_sqrt',
'mwc_tp020_sqrt', 'mcl_tp020_sqrt', 'mcr_tp020_sqrt', 'mm_tp020_sqrt',
'ks_tp020_sqrt', 'total_ms_tp030_sqrt', 'mw_tp030_sqrt',
'mwc_tp030_sqrt', 'mcl_tp030_sqrt', 'mcr_tp030_sqrt', 'mm_tp030_sqrt',
'ks_tp030_sqrt', 'total_ms_tp040_sqrt', 'mw_tp040_sqrt',
'mwc_tp040_sqrt', 'mcl_tp040_sqrt', 'mcr_tp040_sqrt', 'mm_tp040_sqrt',
'ks_tp040_sqrt', 'total_ms_tp050_sqrt', 'mw_tp050_sqrt',
'mwc_tp050_sqrt', 'mcl_tp050_sqrt', 'mcr_tp050_sqrt', 'mm_tp050_sqrt',
'ks_tp050_sqrt', 'total_ms_tp060_sqrt', 'mw_tp060_sqrt',
'mwc_tp060_sqrt', 'mcl_tp060_sqrt', 'mcr_tp060_sqrt', 'mm_tp060_sqrt',
'ks_tp060_sqrt', 'total_ms_tp070_sqrt', 'mw_tp070_sqrt',
'mwc_tp070_sqrt', 'mcl_tp070_sqrt', 'mcr_tp070_sqrt', 'mm_tp070_sqrt',
'ks_tp070_sqrt', 'total_ms_tp080_sqrt', 'mw_tp080_sqrt',
'mwc_tp080_sqrt', 'mcl_tp080_sqrt', 'mcr_tp080_sqrt', 'mm_tp080_sqrt',
'ks_tp080_sqrt', 'total_ms_tp090_sqrt', 'mw_tp090_sqrt',
'mwc_tp090_sqrt', 'mcl_tp090_sqrt', 'mcr_tp090_sqrt', 'mm_tp090_sqrt',
'ks_tp090_sqrt', 'total_ms_tp100_sqrt', 'mw_tp100_sqrt',
'mwc_tp100_sqrt', 'mcl_tp100_sqrt', 'mcr_tp100_sqrt', 'mm_tp100_sqrt',
'ks_tp100_sqrt'],
dtype='object')
additive_transform.fit_transform( select_sess_X ).shape
(684, 106)
enet_to_fit = LogisticRegression(penalty='elasticnet', solver='saga', random_state=101, max_iter=25001, fit_intercept=True)
enet_grid = {'enet__C': np.exp(np.linspace(-6, 6, num=11)),
'enet__l1_ratio': np.linspace(0, 1, num=5)}
enet_additv_wflow = Pipeline( steps = [('prepro', additive_transform),
('enet', enet_to_fit)] )
my_cv = StratifiedKFold(n_splits=5, shuffle=True, random_state=101)
enet_additv_search = GridSearchCV( enet_additv_wflow, enet_grid, cv=my_cv)
enet_additv_search_results = enet_additv_search.fit(select_sess_X, select_sess_y)
enet_additv_search_results.best_params_
{'enet__C': 121.51041751873476, 'enet__l1_ratio': 0.75}
enet_additv_search_results.best_score_
1.0
enet_additv_search_results.best_estimator_.named_steps['enet'].coef_
array([[ -2.99205639, -0.33862077, -1.7724737 , 1.11810899,
-0.46804275, 1.18837941, -3.17587945, 0.42600517,
-0.63108182, -0.07700163, -0.04601613, 0.31609295,
3.27131202, 2.28858518, 0.12927447, 2.26359491,
2.2517737 , 3.78548011, 0.59060774, 4.09351568,
0. , 5.33559167, 6.30823559, 2.18596404,
-7.82398504, 4.29544352, 2.99143548, -6.14081782,
3.43377446, -3.68567434, -1.12506957, -8.93486151,
1.80338888, 2.46869709, 3.11239468, -4.58546883,
-2.38070291, 2.92099139, 4.80392527, -0.81221845,
3.63232456, -3.50409138, -7.89906217, 1.26223988,
-2.23171446, -2.66488168, 0.05358037, -1.10539039,
-2.21432191, -5.63029727, -11.93313654, -1.79135305,
-2.16174647, 1.97475301, 3.11895145, 6.34528489,
4.41134441, 5.66732412, -3.46016746, -4.23059532,
-4.39070288, -3.85256465, 3.04650392, 1.83311483,
-0.7715605 , 5.06976793, 11.72527673, 0.34841625,
-6.71684482, 3.5537281 , 0.05580302, 5.45714064,
3.99725446, 0. , 0. , -5.32439665,
-9.14346183, -4.37105834, 3.33363713, 2.30458718,
5.81897268, 1.26223584, -0.38332789, -3.53805586,
3.91202351, -3.57065593, 1.0985811 , 2.77637185,
0.72782452, 2.43770301, 8.98005868, -7.87336199,
-7.84850724, -2.14168934, 2.23844561, -0.37718957,
0. , 0.69844083, 0.67404907, 0.89081033,
0.40344222, 0.57167265, 1.38095821, 0.77793707,
0.1290023 , 0.59332223]])
enet_additv_search_results.best_estimator_.named_steps['enet'].coef_.shape
(1, 106)
enet_additv_search_results.best_estimator_.named_steps['enet'].coef_.tolist()[0][85]
-3.570655932050126
enet_additv_search_results.best_estimator_.named_steps['enet'].coef_[enet_additv_search_results.best_estimator_.named_steps['enet'].coef_ != 0].size
102
select_sess_df.shape
(684, 81)
select_sess_df.columns
Index(['sess', 'sid', 'actv_grp', 'total_ms_tp000_sqrt', 'mw_tp000_sqrt',
'mwc_tp000_sqrt', 'mcl_tp000_sqrt', 'mcr_tp000_sqrt', 'mm_tp000_sqrt',
'ks_tp000_sqrt', 'total_ms_tp010_sqrt', 'mw_tp010_sqrt',
'mwc_tp010_sqrt', 'mcl_tp010_sqrt', 'mcr_tp010_sqrt', 'mm_tp010_sqrt',
'ks_tp010_sqrt', 'total_ms_tp020_sqrt', 'mw_tp020_sqrt',
'mwc_tp020_sqrt', 'mcl_tp020_sqrt', 'mcr_tp020_sqrt', 'mm_tp020_sqrt',
'ks_tp020_sqrt', 'total_ms_tp030_sqrt', 'mw_tp030_sqrt',
'mwc_tp030_sqrt', 'mcl_tp030_sqrt', 'mcr_tp030_sqrt', 'mm_tp030_sqrt',
'ks_tp030_sqrt', 'total_ms_tp040_sqrt', 'mw_tp040_sqrt',
'mwc_tp040_sqrt', 'mcl_tp040_sqrt', 'mcr_tp040_sqrt', 'mm_tp040_sqrt',
'ks_tp040_sqrt', 'total_ms_tp050_sqrt', 'mw_tp050_sqrt',
'mwc_tp050_sqrt', 'mcl_tp050_sqrt', 'mcr_tp050_sqrt', 'mm_tp050_sqrt',
'ks_tp050_sqrt', 'total_ms_tp060_sqrt', 'mw_tp060_sqrt',
'mwc_tp060_sqrt', 'mcl_tp060_sqrt', 'mcr_tp060_sqrt', 'mm_tp060_sqrt',
'ks_tp060_sqrt', 'total_ms_tp070_sqrt', 'mw_tp070_sqrt',
'mwc_tp070_sqrt', 'mcl_tp070_sqrt', 'mcr_tp070_sqrt', 'mm_tp070_sqrt',
'ks_tp070_sqrt', 'total_ms_tp080_sqrt', 'mw_tp080_sqrt',
'mwc_tp080_sqrt', 'mcl_tp080_sqrt', 'mcr_tp080_sqrt', 'mm_tp080_sqrt',
'ks_tp080_sqrt', 'total_ms_tp090_sqrt', 'mw_tp090_sqrt',
'mwc_tp090_sqrt', 'mcl_tp090_sqrt', 'mcr_tp090_sqrt', 'mm_tp090_sqrt',
'ks_tp090_sqrt', 'total_ms_tp100_sqrt', 'mw_tp100_sqrt',
'mwc_tp100_sqrt', 'mcl_tp100_sqrt', 'mcr_tp100_sqrt', 'mm_tp100_sqrt',
'ks_tp100_sqrt', 'interim_pass'],
dtype='object')
fig, ax = plt.subplots(figsize=(24, 12))
sns.heatmap( data = np.transpose( calc_pca_contrib( pca_object.components_) ) > (1 / timepoint_pcs.shape[1]),
xticklabels=pc_scores_df.iloc[:, :len(feature_names)].columns.tolist(),
yticklabels=feature_names,
vmin=0, vmax=1)
plt.savefig('s3_pc_contrib.png')
plt.show()
select_sess_df.sid.unique()
array([ 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15,
16, 17, 18, 19, 20, 23, 24, 25, 27, 28, 29, 30, 31,
32, 34, 35, 36, 38, 39, 41, 42, 43, 44, 45, 47, 48,
49, 50, 51, 52, 53, 54, 55, 56, 59, 61, 63, 66, 67,
68, 69, 70, 72, 73, 74, 75, 76, 78, 79, 80, 81, 82,
83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95,
97, 98, 99, 100, 101, 102, 105, 106])
select_sess_df.actv_grp.unique()
array(['Aulaweb', 'Blank', 'Deeds', 'Diagram', 'Other', 'Properties',
'Study', 'TextEditor', 'FSM_Related', 'Study_Materials'],
dtype=object)
input_grid = pd.DataFrame([ (xs,xag,xa0,xb0,xc0,xd0,xe0,xf0,xg0,xa1,xb1,xc1,xd1,xe1,xf1,xg1,xa2,xb2,xc2,xd2,xe2,xf2,xg2,xa3,xb3,xc3,xd3,xe3,xf3,xg3,xa4,xb4,xc4,xd4,xe4,xf4,xg4,xa5,xb5,xc5,xd5,xe5,xf5,xg5,xa6,xb6,xc6,xd6,xe6,xf6,xg6,xa7,xb7,xc7,xd7,xe7,xf7,xg7,xa8,xb8,xc8,xd8,xe8,xf8,xg8,xa9,xb9,xc9,xd9,xe9,xf9,xg9,xa10,xb10,xc10,xd10,xe10,xf10,xg10)
for xs in [5,20,41,56,79,93]
for xag in ['Aulaweb', 'Deeds', 'Diagram','TextEditor']
for xa0 in [select_sess_df.total_ms_tp000_sqrt.median()]
for xb0 in [select_sess_df.mw_tp000_sqrt.median()]
for xc0 in [select_sess_df.mwc_tp000_sqrt.median()]
for xd0 in [select_sess_df.mcl_tp000_sqrt.median()]
for xe0 in [select_sess_df.mcr_tp000_sqrt.median()]
for xf0 in [select_sess_df.mm_tp000_sqrt.median()]
for xg0 in [select_sess_df.ks_tp000_sqrt.median()]
for xa1 in [select_sess_df.total_ms_tp010_sqrt.median()]
for xb1 in [select_sess_df.mw_tp010_sqrt.median()]
for xc1 in [select_sess_df.mwc_tp010_sqrt.median()]
for xd1 in [select_sess_df.mcl_tp010_sqrt.median()]
for xe1 in [select_sess_df.mcr_tp010_sqrt.median()]
for xf1 in [select_sess_df.mm_tp010_sqrt.median()]
for xg1 in [select_sess_df.ks_tp010_sqrt.median()]
for xa2 in [select_sess_df.total_ms_tp020_sqrt.median()]
for xb2 in [select_sess_df.mw_tp020_sqrt.median()]
for xc2 in [select_sess_df.mwc_tp020_sqrt.median()]
for xd2 in [select_sess_df.mcl_tp020_sqrt.median()]
for xe2 in [select_sess_df.mcr_tp020_sqrt.median()]
for xf2 in [select_sess_df.mm_tp020_sqrt.median()]
for xg2 in [select_sess_df.ks_tp020_sqrt.median()]
for xa3 in [select_sess_df.total_ms_tp030_sqrt.median()]
for xb3 in [select_sess_df.mw_tp030_sqrt.median()]
for xc3 in [select_sess_df.mwc_tp030_sqrt.median()]
for xd3 in [select_sess_df.mcl_tp030_sqrt.median()]
for xe3 in [select_sess_df.mcr_tp030_sqrt.median()]
for xf3 in [select_sess_df.mm_tp030_sqrt.median()]
for xg3 in [select_sess_df.ks_tp030_sqrt.median()]
for xa4 in [select_sess_df.total_ms_tp040_sqrt.median()]
for xb4 in [select_sess_df.mw_tp040_sqrt.median()]
for xc4 in [select_sess_df.mwc_tp040_sqrt.median()]
for xd4 in [select_sess_df.mcl_tp040_sqrt.median()]
for xe4 in [select_sess_df.mcr_tp040_sqrt.median()]
for xf4 in [select_sess_df.mm_tp040_sqrt.median()]
for xg4 in [select_sess_df.ks_tp040_sqrt.median()]
for xa5 in [select_sess_df.total_ms_tp050_sqrt.median()]
for xb5 in [select_sess_df.mw_tp050_sqrt.median()]
for xc5 in [select_sess_df.mwc_tp050_sqrt.median()]
for xd5 in [select_sess_df.mcl_tp050_sqrt.median()]
for xe5 in [select_sess_df.mcr_tp050_sqrt.median()]
for xf5 in [select_sess_df.mm_tp050_sqrt.median()]
for xg5 in [select_sess_df.ks_tp050_sqrt.median()]
for xa6 in [select_sess_df.total_ms_tp060_sqrt.median()]
for xb6 in [select_sess_df.mw_tp060_sqrt.median()]
for xc6 in [select_sess_df.mwc_tp060_sqrt.median()]
for xd6 in [select_sess_df.mcl_tp060_sqrt.median()]
for xe6 in [select_sess_df.mcr_tp060_sqrt.median()]
for xf6 in [select_sess_df.mm_tp060_sqrt.median()]
for xg6 in [select_sess_df.ks_tp060_sqrt.median()]
for xa7 in [select_sess_df.total_ms_tp070_sqrt.median()]
for xb7 in [select_sess_df.mw_tp070_sqrt.median()]
for xc7 in [select_sess_df.mwc_tp070_sqrt.median()]
for xd7 in [select_sess_df.mcl_tp070_sqrt.median()]
for xe7 in [select_sess_df.mcr_tp070_sqrt.median()]
for xf7 in [select_sess_df.mm_tp070_sqrt.median()]
for xg7 in [select_sess_df.ks_tp070_sqrt.median()]
for xa8 in [select_sess_df.total_ms_tp080_sqrt.median()]
for xb8 in [select_sess_df.mw_tp080_sqrt.median()]
for xc8 in [select_sess_df.mwc_tp080_sqrt.median()]
for xd8 in [select_sess_df.mcl_tp080_sqrt.median()]
for xe8 in [select_sess_df.mcr_tp080_sqrt.median()]
for xf8 in [select_sess_df.mm_tp080_sqrt.median()]
for xg8 in [select_sess_df.ks_tp080_sqrt.median()]
for xa9 in [select_sess_df.total_ms_tp090_sqrt.median()]
for xb9 in [select_sess_df.mw_tp090_sqrt.median()]
for xc9 in [select_sess_df.mwc_tp090_sqrt.median()]
for xd9 in [select_sess_df.mcl_tp090_sqrt.median()]
for xe9 in [select_sess_df.mcr_tp090_sqrt.median()]
for xf9 in [select_sess_df.mm_tp090_sqrt.median()]
for xg9 in [select_sess_df.ks_tp090_sqrt.median()]
for xa10 in [select_sess_df.total_ms_tp100_sqrt.median()]
for xb10 in [select_sess_df.mw_tp100_sqrt.median()]
for xc10 in [select_sess_df.mwc_tp100_sqrt.median()]
for xd10 in [select_sess_df.mcl_tp100_sqrt.median()]
for xe10 in [select_sess_df.mcr_tp100_sqrt.median()]
for xf10 in np.linspace(select_sess_df.mm_tp100_sqrt.min() - 0.02, select_sess_df.mm_tp100_sqrt.max() + 0.02, num=101)
for xg10 in [select_sess_df.ks_tp100_sqrt.median()]],
columns=['sid', 'actv_grp', 'total_ms_tp000_sqrt', 'mw_tp000_sqrt',
'mwc_tp000_sqrt', 'mcl_tp000_sqrt', 'mcr_tp000_sqrt', 'mm_tp000_sqrt',
'ks_tp000_sqrt', 'total_ms_tp010_sqrt', 'mw_tp010_sqrt',
'mwc_tp010_sqrt', 'mcl_tp010_sqrt', 'mcr_tp010_sqrt', 'mm_tp010_sqrt',
'ks_tp010_sqrt', 'total_ms_tp020_sqrt', 'mw_tp020_sqrt',
'mwc_tp020_sqrt', 'mcl_tp020_sqrt', 'mcr_tp020_sqrt', 'mm_tp020_sqrt',
'ks_tp020_sqrt', 'total_ms_tp030_sqrt', 'mw_tp030_sqrt',
'mwc_tp030_sqrt', 'mcl_tp030_sqrt', 'mcr_tp030_sqrt', 'mm_tp030_sqrt',
'ks_tp030_sqrt', 'total_ms_tp040_sqrt', 'mw_tp040_sqrt',
'mwc_tp040_sqrt', 'mcl_tp040_sqrt', 'mcr_tp040_sqrt', 'mm_tp040_sqrt',
'ks_tp040_sqrt', 'total_ms_tp050_sqrt', 'mw_tp050_sqrt',
'mwc_tp050_sqrt', 'mcl_tp050_sqrt', 'mcr_tp050_sqrt', 'mm_tp050_sqrt',
'ks_tp050_sqrt', 'total_ms_tp060_sqrt', 'mw_tp060_sqrt',
'mwc_tp060_sqrt', 'mcl_tp060_sqrt', 'mcr_tp060_sqrt', 'mm_tp060_sqrt',
'ks_tp060_sqrt', 'total_ms_tp070_sqrt', 'mw_tp070_sqrt',
'mwc_tp070_sqrt', 'mcl_tp070_sqrt', 'mcr_tp070_sqrt', 'mm_tp070_sqrt',
'ks_tp070_sqrt', 'total_ms_tp080_sqrt', 'mw_tp080_sqrt',
'mwc_tp080_sqrt', 'mcl_tp080_sqrt', 'mcr_tp080_sqrt', 'mm_tp080_sqrt',
'ks_tp080_sqrt', 'total_ms_tp090_sqrt', 'mw_tp090_sqrt',
'mwc_tp090_sqrt', 'mcl_tp090_sqrt', 'mcr_tp090_sqrt', 'mm_tp090_sqrt',
'ks_tp090_sqrt', 'total_ms_tp100_sqrt', 'mw_tp100_sqrt',
'mwc_tp100_sqrt', 'mcl_tp100_sqrt', 'mcr_tp100_sqrt', 'mm_tp100_sqrt',
'ks_tp100_sqrt'])
input_grid.shape
(2424, 79)
input_grid.head()
| sid | actv_grp | total_ms_tp000_sqrt | mw_tp000_sqrt | mwc_tp000_sqrt | mcl_tp000_sqrt | mcr_tp000_sqrt | mm_tp000_sqrt | ks_tp000_sqrt | total_ms_tp010_sqrt | ... | mcr_tp090_sqrt | mm_tp090_sqrt | ks_tp090_sqrt | total_ms_tp100_sqrt | mw_tp100_sqrt | mwc_tp100_sqrt | mcl_tp100_sqrt | mcr_tp100_sqrt | mm_tp100_sqrt | ks_tp100_sqrt | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 5 | Aulaweb | 406.20192 | 0.0 | 0.0 | 6.402648 | 0.0 | 55.063383 | 7.211103 | 1533.622944 | ... | 17.320508 | 411.197016 | 63.027771 | 3042.778281 | 45.977966 | 0.0 | 61.749493 | 17.720045 | 24.779194 | 65.061509 |
| 1 | 5 | Aulaweb | 406.20192 | 0.0 | 0.0 | 6.402648 | 0.0 | 55.063383 | 7.211103 | 1533.622944 | ... | 17.320508 | 411.197016 | 63.027771 | 3042.778281 | 45.977966 | 0.0 | 61.749493 | 17.720045 | 34.331949 | 65.061509 |
| 2 | 5 | Aulaweb | 406.20192 | 0.0 | 0.0 | 6.402648 | 0.0 | 55.063383 | 7.211103 | 1533.622944 | ... | 17.320508 | 411.197016 | 63.027771 | 3042.778281 | 45.977966 | 0.0 | 61.749493 | 17.720045 | 43.884704 | 65.061509 |
| 3 | 5 | Aulaweb | 406.20192 | 0.0 | 0.0 | 6.402648 | 0.0 | 55.063383 | 7.211103 | 1533.622944 | ... | 17.320508 | 411.197016 | 63.027771 | 3042.778281 | 45.977966 | 0.0 | 61.749493 | 17.720045 | 53.437459 | 65.061509 |
| 4 | 5 | Aulaweb | 406.20192 | 0.0 | 0.0 | 6.402648 | 0.0 | 55.063383 | 7.211103 | 1533.622944 | ... | 17.320508 | 411.197016 | 63.027771 | 3042.778281 | 45.977966 | 0.0 | 61.749493 | 17.720045 | 62.990214 | 65.061509 |
5 rows × 79 columns
select_sess_X.head()
| sid | actv_grp | total_ms_tp000_sqrt | mw_tp000_sqrt | mwc_tp000_sqrt | mcl_tp000_sqrt | mcr_tp000_sqrt | mm_tp000_sqrt | ks_tp000_sqrt | total_ms_tp010_sqrt | ... | mcr_tp090_sqrt | mm_tp090_sqrt | ks_tp090_sqrt | total_ms_tp100_sqrt | mw_tp100_sqrt | mwc_tp100_sqrt | mcl_tp100_sqrt | mcr_tp100_sqrt | mm_tp100_sqrt | ks_tp100_sqrt | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 653 | 2 | Aulaweb | 2059.368835 | 14.628739 | 0.0 | 42.166337 | 10.862780 | 324.242193 | 41.121770 | 2066.784943 | ... | 17.175564 | 505.788197 | 67.339439 | 3129.376935 | 37.134889 | 0.0 | 75.139870 | 17.435596 | 514.693112 | 69.806876 |
| 654 | 2 | Blank | 89.442719 | 0.000000 | 0.0 | 2.000000 | 0.000000 | 22.271057 | 0.000000 | 741.080293 | ... | 17.146428 | 504.451385 | 67.059675 | 3107.893177 | 35.958309 | 0.0 | 74.565408 | 17.262677 | 509.718550 | 69.123079 |
| 655 | 2 | Deeds | 187.082869 | 2.828427 | 0.0 | 3.464102 | 0.000000 | 33.570821 | 0.000000 | 1090.321054 | ... | 17.146428 | 506.464411 | 67.652051 | 3129.057366 | 37.134889 | 0.0 | 75.113248 | 17.435596 | 514.478377 | 69.806876 |
| 656 | 2 | Diagram | 2018.910597 | 14.560220 | 0.0 | 41.448764 | 10.488088 | 316.633858 | 36.510273 | 2056.404629 | ... | 17.378147 | 512.338755 | 69.778220 | 3129.057366 | 37.134889 | 0.0 | 75.099933 | 17.435596 | 514.401594 | 69.806876 |
| 657 | 2 | Other | 70.710678 | 0.000000 | 0.0 | 1.414214 | 0.000000 | 16.248077 | 0.000000 | 1159.353268 | ... | 17.492856 | 515.346679 | 70.289402 | 3135.442552 | 37.134889 | 0.0 | 75.246262 | 17.492856 | 515.694677 | 70.349129 |
5 rows × 79 columns
input_grid_copy = input_grid.copy()
input_grid_copy['pred_enet_additv_class'] = enet_additv_search_results.predict( input_grid )
sns.relplot(data = input_grid_copy, x='mm_tp100_sqrt', y='pred_enet_additv_class', hue='actv_grp',
col='sid', col_wrap=2, kind='line')
plt.show()
categorical_inputs_b = ['actv_grp']
categorical_transform_b = Pipeline( steps = [ ('dummy', OneHotEncoder(drop='first')) ] )
Use numeric_transform_pca from above
complex_transform_b = ColumnTransformer( transformers=[('num', numeric_transform_pca, sqrt_vars),
('cat', categorical_transform_b, categorical_inputs_b)])
complex_transform_b.transformers
[('num',
Pipeline(steps=[('std_input', StandardScaler()), ('pca', PCA(n_components=12)),
('std_features', StandardScaler())]),
['total_ms_tp000_sqrt',
'mw_tp000_sqrt',
'mwc_tp000_sqrt',
'mcl_tp000_sqrt',
'mcr_tp000_sqrt',
'mm_tp000_sqrt',
'ks_tp000_sqrt',
'total_ms_tp010_sqrt',
'mw_tp010_sqrt',
'mwc_tp010_sqrt',
'mcl_tp010_sqrt',
'mcr_tp010_sqrt',
'mm_tp010_sqrt',
'ks_tp010_sqrt',
'total_ms_tp020_sqrt',
'mw_tp020_sqrt',
'mwc_tp020_sqrt',
'mcl_tp020_sqrt',
'mcr_tp020_sqrt',
'mm_tp020_sqrt',
'ks_tp020_sqrt',
'total_ms_tp030_sqrt',
'mw_tp030_sqrt',
'mwc_tp030_sqrt',
'mcl_tp030_sqrt',
'mcr_tp030_sqrt',
'mm_tp030_sqrt',
'ks_tp030_sqrt',
'total_ms_tp040_sqrt',
'mw_tp040_sqrt',
'mwc_tp040_sqrt',
'mcl_tp040_sqrt',
'mcr_tp040_sqrt',
'mm_tp040_sqrt',
'ks_tp040_sqrt',
'total_ms_tp050_sqrt',
'mw_tp050_sqrt',
'mwc_tp050_sqrt',
'mcl_tp050_sqrt',
'mcr_tp050_sqrt',
'mm_tp050_sqrt',
'ks_tp050_sqrt',
'total_ms_tp060_sqrt',
'mw_tp060_sqrt',
'mwc_tp060_sqrt',
'mcl_tp060_sqrt',
'mcr_tp060_sqrt',
'mm_tp060_sqrt',
'ks_tp060_sqrt',
'total_ms_tp070_sqrt',
'mw_tp070_sqrt',
'mwc_tp070_sqrt',
'mcl_tp070_sqrt',
'mcr_tp070_sqrt',
'mm_tp070_sqrt',
'ks_tp070_sqrt',
'total_ms_tp080_sqrt',
'mw_tp080_sqrt',
'mwc_tp080_sqrt',
'mcl_tp080_sqrt',
'mcr_tp080_sqrt',
'mm_tp080_sqrt',
'ks_tp080_sqrt',
'total_ms_tp090_sqrt',
'mw_tp090_sqrt',
'mwc_tp090_sqrt',
'mcl_tp090_sqrt',
'mcr_tp090_sqrt',
'mm_tp090_sqrt',
'ks_tp090_sqrt',
'total_ms_tp100_sqrt',
'mw_tp100_sqrt',
'mwc_tp100_sqrt',
'mcl_tp100_sqrt',
'mcr_tp100_sqrt',
'mm_tp100_sqrt',
'ks_tp100_sqrt']),
('cat',
Pipeline(steps=[('dummy', OneHotEncoder(drop='first'))]),
['actv_grp'])]
interact_transform = Pipeline( steps=[ ('prepro', complex_transform_b),
('make_poly', PolynomialFeatures(degree=2, interaction_only=True, include_bias=False))])
interact_transform.fit_transform(select_sess_df).shape
(684, 231)
select_sess_X_b = select_sess_df.drop(columns=['sess','sid','interim_pass']).copy()
select_sess_X.shape
(684, 79)
select_sess_X_b.shape
(684, 78)
select_sess_X_b.actv_grp.unique()
array(['Aulaweb', 'Blank', 'Deeds', 'Diagram', 'Other', 'Properties',
'Study', 'TextEditor', 'FSM_Related', 'Study_Materials'],
dtype=object)
len(select_sess_X_b.actv_grp.unique())
10
select_sess_df.interim_pass.value_counts()
1.0 399 0.0 285 Name: interim_pass, dtype: int64
Use same select_sess_y as above
interact_transform.fit_transform( select_sess_X_b ).shape
(684, 231)
Use enet_to_fit, enet_grid, and my_cv from above
enet_interact_wflow = Pipeline( steps = [('prepro', interact_transform),
('enet', enet_to_fit)] )
enet_interact_search = GridSearchCV( enet_interact_wflow, enet_grid, cv=my_cv)
enet_interact_search_results = enet_interact_search.fit(select_sess_X_b, select_sess_y)
enet_interact_search_results.best_params_
{'enet__C': 0.09071795328941247, 'enet__l1_ratio': 0.25}
enet_interact_search_results.best_score_
0.852361528553027
enet_interact_search_results.best_estimator_.named_steps['enet'].coef_
array([[-0.99568765, -0.00117056, -0.60305759, 0.27520886, -0.0533723 ,
0.22949601, -0.90479439, 0.01257915, -0.18802707, 0.19718861,
-0.27377263, 0.29851809, 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , -0.06123962, 0.26603149, -0.31558159, -0.03139083,
-0.26918984, -0.393452 , 0.20288697, 0.33194558, -0.27483153,
0.1313263 , 0. , 0. , 0. , -0.05512202,
0. , 0. , 0.04287527, -0.20251278, 0. ,
0. , 0.21296741, 0.29859898, 0.15593914, 0.01785055,
0.014245 , 0. , 0. , 0. , -0.08858999,
0. , 0. , 0. , 0. , 0. ,
0. , 0.01477685, -0.00900938, 0. , 0. ,
0.14779542, 0.15749979, -0.33605648, -0.06916644, 0. ,
0. , 0.16302192, 0.14095135, -0.02570328, 0. ,
0. , 0. , 0. , 0. , -0.07801566,
0. , 0. , 0. , 0. , -0.03419471,
0.43802044, -0.0043646 , 0.10536862, 0. , 0.2355448 ,
0.11054104, 0. , 0. , 0. , 0. ,
0. , 0.0227873 , 0. , 0. , 0. ,
0. , -0.34334666, 0.09063219, -0.16927434, -0.17174628,
0. , -0.08179283, 0. , 0. , 0. ,
0. , 0. , -0.04115672, 0. , 0. ,
0. , -0.04611465, -0.0448777 , 0.35125935, 0.05664544,
0.07382522, -0.1310395 , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , -0.13259513, -0.05504354, 0. , 0. ,
0.04513034, 0. , 0. , 0. , 0. ,
0. , -0.33080762, 0.09594799, 0. , 0. ,
0.03232009, 0.07344064, -0.05983068, 0.09352574, 0. ,
-0.09779666, 0.48175805, 0. , 0. , 0. ,
0.26332381, 0. , 0. , -0.20481506, 0. ,
0.25483925, 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0.23500379, -0.11617069, -0.20363709, 0. , 0.0240947 ,
0. , 0. , 0. , 0. , 0. ,
0. , -0.06644953, 0. , -0.15026159, 0.06279823,
0. , 0. , 0. , -0.12581009, 0. ,
-0.03275198, 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 0. ,
0. ]])
enet_interact_search_results.best_estimator_.named_steps['enet'].coef_.shape
(1, 231)
enet_interact_search_results.best_estimator_.named_steps['enet'].coef_[enet_interact_search_results.best_estimator_.named_steps['enet'].coef_ != 0].size
83
input_grid_b = pd.DataFrame([ (xag,xa0,xb0,xc0,xd0,xe0,xf0,xg0,xa1,xb1,xc1,xd1,xe1,xf1,xg1,xa2,xb2,xc2,xd2,xe2,xf2,xg2,xa3,xb3,xc3,xd3,xe3,xf3,xg3,xa4,xb4,xc4,xd4,xe4,xf4,xg4,xa5,xb5,xc5,xd5,xe5,xf5,xg5,xa6,xb6,xc6,xd6,xe6,xf6,xg6,xa7,xb7,xc7,xd7,xe7,xf7,xg7,xa8,xb8,xc8,xd8,xe8,xf8,xg8,xa9,xb9,xc9,xd9,xe9,xf9,xg9,xa10,xb10,xc10,xd10,xe10,xf10,xg10)
for xag in select_sess_df.actv_grp.unique()
for xa0 in [select_sess_df.total_ms_tp000_sqrt.median()]
for xb0 in [select_sess_df.mw_tp000_sqrt.median()]
for xc0 in [select_sess_df.mwc_tp000_sqrt.median()]
for xd0 in [select_sess_df.mcl_tp000_sqrt.median()]
for xe0 in [select_sess_df.mcr_tp000_sqrt.median()]
for xf0 in [select_sess_df.mm_tp000_sqrt.median()]
for xg0 in [select_sess_df.ks_tp000_sqrt.median()]
for xa1 in [select_sess_df.total_ms_tp010_sqrt.median()]
for xb1 in [select_sess_df.mw_tp010_sqrt.median()]
for xc1 in [select_sess_df.mwc_tp010_sqrt.median()]
for xd1 in [select_sess_df.mcl_tp010_sqrt.median()]
for xe1 in [select_sess_df.mcr_tp010_sqrt.median()]
for xf1 in [select_sess_df.mm_tp010_sqrt.median()]
for xg1 in [select_sess_df.ks_tp010_sqrt.median()]
for xa2 in [select_sess_df.total_ms_tp020_sqrt.median()]
for xb2 in [select_sess_df.mw_tp020_sqrt.median()]
for xc2 in [select_sess_df.mwc_tp020_sqrt.median()]
for xd2 in [select_sess_df.mcl_tp020_sqrt.median()]
for xe2 in [select_sess_df.mcr_tp020_sqrt.median()]
for xf2 in [select_sess_df.mm_tp020_sqrt.median()]
for xg2 in [select_sess_df.ks_tp020_sqrt.median()]
for xa3 in [select_sess_df.total_ms_tp030_sqrt.median()]
for xb3 in [select_sess_df.mw_tp030_sqrt.median()]
for xc3 in [select_sess_df.mwc_tp030_sqrt.median()]
for xd3 in [select_sess_df.mcl_tp030_sqrt.median()]
for xe3 in [select_sess_df.mcr_tp030_sqrt.median()]
for xf3 in [select_sess_df.mm_tp030_sqrt.median()]
for xg3 in [select_sess_df.ks_tp030_sqrt.median()]
for xa4 in [select_sess_df.total_ms_tp040_sqrt.median()]
for xb4 in [select_sess_df.mw_tp040_sqrt.median()]
for xc4 in [select_sess_df.mwc_tp040_sqrt.median()]
for xd4 in [select_sess_df.mcl_tp040_sqrt.median()]
for xe4 in [select_sess_df.mcr_tp040_sqrt.median()]
for xf4 in [select_sess_df.mm_tp040_sqrt.median()]
for xg4 in [select_sess_df.ks_tp040_sqrt.median()]
for xa5 in [select_sess_df.total_ms_tp050_sqrt.median()]
for xb5 in [select_sess_df.mw_tp050_sqrt.median()]
for xc5 in [select_sess_df.mwc_tp050_sqrt.median()]
for xd5 in [select_sess_df.mcl_tp050_sqrt.median()]
for xe5 in [select_sess_df.mcr_tp050_sqrt.median()]
for xf5 in [select_sess_df.mm_tp050_sqrt.median()]
for xg5 in [select_sess_df.ks_tp050_sqrt.median()]
for xa6 in [select_sess_df.total_ms_tp060_sqrt.median()]
for xb6 in [select_sess_df.mw_tp060_sqrt.median()]
for xc6 in [select_sess_df.mwc_tp060_sqrt.median()]
for xd6 in [select_sess_df.mcl_tp060_sqrt.median()]
for xe6 in [select_sess_df.mcr_tp060_sqrt.median()]
for xf6 in [select_sess_df.mm_tp060_sqrt.median()]
for xg6 in [select_sess_df.ks_tp060_sqrt.median()]
for xa7 in [select_sess_df.total_ms_tp070_sqrt.median()]
for xb7 in [select_sess_df.mw_tp070_sqrt.median()]
for xc7 in [select_sess_df.mwc_tp070_sqrt.median()]
for xd7 in [select_sess_df.mcl_tp070_sqrt.median()]
for xe7 in [select_sess_df.mcr_tp070_sqrt.median()]
for xf7 in [select_sess_df.mm_tp070_sqrt.median()]
for xg7 in [select_sess_df.ks_tp070_sqrt.median()]
for xa8 in [select_sess_df.total_ms_tp080_sqrt.median()]
for xb8 in [select_sess_df.mw_tp080_sqrt.median()]
for xc8 in [select_sess_df.mwc_tp080_sqrt.median()]
for xd8 in [select_sess_df.mcl_tp080_sqrt.median()]
for xe8 in [select_sess_df.mcr_tp080_sqrt.median()]
for xf8 in [select_sess_df.mm_tp080_sqrt.median()]
for xg8 in [select_sess_df.ks_tp080_sqrt.median()]
for xa9 in [select_sess_df.total_ms_tp090_sqrt.median()]
for xb9 in [select_sess_df.mw_tp090_sqrt.median()]
for xc9 in [select_sess_df.mwc_tp090_sqrt.median()]
for xd9 in [select_sess_df.mcl_tp090_sqrt.median()]
for xe9 in [select_sess_df.mcr_tp090_sqrt.median()]
for xf9 in [select_sess_df.mm_tp090_sqrt.median()]
for xg9 in [select_sess_df.ks_tp090_sqrt.median()]
for xa10 in [select_sess_df.total_ms_tp100_sqrt.median()]
for xb10 in [select_sess_df.mw_tp100_sqrt.median()]
for xc10 in [select_sess_df.mwc_tp100_sqrt.median()]
for xd10 in [select_sess_df.mcl_tp100_sqrt.median()]
for xe10 in [select_sess_df.mcr_tp100_sqrt.median()]
for xf10 in np.linspace(select_sess_df.mm_tp100_sqrt.min() - 0.02, select_sess_df.mm_tp100_sqrt.max() + 0.02, num=101)
for xg10 in [select_sess_df.ks_tp100_sqrt.median()]],
columns=['actv_grp', 'total_ms_tp000_sqrt', 'mw_tp000_sqrt',
'mwc_tp000_sqrt', 'mcl_tp000_sqrt', 'mcr_tp000_sqrt', 'mm_tp000_sqrt',
'ks_tp000_sqrt', 'total_ms_tp010_sqrt', 'mw_tp010_sqrt',
'mwc_tp010_sqrt', 'mcl_tp010_sqrt', 'mcr_tp010_sqrt', 'mm_tp010_sqrt',
'ks_tp010_sqrt', 'total_ms_tp020_sqrt', 'mw_tp020_sqrt',
'mwc_tp020_sqrt', 'mcl_tp020_sqrt', 'mcr_tp020_sqrt', 'mm_tp020_sqrt',
'ks_tp020_sqrt', 'total_ms_tp030_sqrt', 'mw_tp030_sqrt',
'mwc_tp030_sqrt', 'mcl_tp030_sqrt', 'mcr_tp030_sqrt', 'mm_tp030_sqrt',
'ks_tp030_sqrt', 'total_ms_tp040_sqrt', 'mw_tp040_sqrt',
'mwc_tp040_sqrt', 'mcl_tp040_sqrt', 'mcr_tp040_sqrt', 'mm_tp040_sqrt',
'ks_tp040_sqrt', 'total_ms_tp050_sqrt', 'mw_tp050_sqrt',
'mwc_tp050_sqrt', 'mcl_tp050_sqrt', 'mcr_tp050_sqrt', 'mm_tp050_sqrt',
'ks_tp050_sqrt', 'total_ms_tp060_sqrt', 'mw_tp060_sqrt',
'mwc_tp060_sqrt', 'mcl_tp060_sqrt', 'mcr_tp060_sqrt', 'mm_tp060_sqrt',
'ks_tp060_sqrt', 'total_ms_tp070_sqrt', 'mw_tp070_sqrt',
'mwc_tp070_sqrt', 'mcl_tp070_sqrt', 'mcr_tp070_sqrt', 'mm_tp070_sqrt',
'ks_tp070_sqrt', 'total_ms_tp080_sqrt', 'mw_tp080_sqrt',
'mwc_tp080_sqrt', 'mcl_tp080_sqrt', 'mcr_tp080_sqrt', 'mm_tp080_sqrt',
'ks_tp080_sqrt', 'total_ms_tp090_sqrt', 'mw_tp090_sqrt',
'mwc_tp090_sqrt', 'mcl_tp090_sqrt', 'mcr_tp090_sqrt', 'mm_tp090_sqrt',
'ks_tp090_sqrt', 'total_ms_tp100_sqrt', 'mw_tp100_sqrt',
'mwc_tp100_sqrt', 'mcl_tp100_sqrt', 'mcr_tp100_sqrt', 'mm_tp100_sqrt',
'ks_tp100_sqrt'])
input_grid_b_copy = input_grid_b.copy()
input_grid_b_copy['pred_enet_interact_class'] = enet_interact_search_results.predict( input_grid_b )
sns.relplot(data = input_grid_b_copy, x='mm_tp100_sqrt', y='pred_enet_interact_class',
col='actv_grp', col_wrap=2, hue='actv_grp', kind='line')
plt.show()
s3_pc_scores_12_df = pc_scores_12_df.copy()
%store s3_pc_scores_12_df
Stored 's3_pc_scores_12_df' (DataFrame)
s3_pc_scores_outp_df = pc_scores_outp_df.copy()
%store s3_pc_scores_outp_df
Stored 's3_pc_scores_outp_df' (DataFrame)
s3_additv_sid_model_coef = enet_additv_search_results.best_estimator_.named_steps['enet'].coef_
%store s3_additv_sid_model_coef
Stored 's3_additv_sid_model_coef' (ndarray)
s3_interact_nosid_model_coef = enet_interact_search_results.best_estimator_.named_steps['enet'].coef_
%store s3_interact_nosid_model_coef
Stored 's3_interact_nosid_model_coef' (ndarray)
s3_additv_sid_model_params = enet_additv_search_results.best_params_
%store s3_additv_sid_model_params
Stored 's3_additv_sid_model_params' (dict)
s3_interact_nosid_model_params = enet_interact_search_results.best_params_
%store s3_interact_nosid_model_params
Stored 's3_interact_nosid_model_params' (dict)
s3_additv_sid_model_score = enet_additv_search_results.best_score_
%store s3_additv_sid_model_score
Stored 's3_additv_sid_model_score' (float64)
s3_interact_nosid_model_score = enet_interact_search_results.best_score_
%store s3_interact_nosid_model_score
Stored 's3_interact_nosid_model_score' (float64)
s3_input_grid_copy = input_grid_copy.copy()
%store s3_input_grid_copy
Stored 's3_input_grid_copy' (DataFrame)
s3_input_grid_b_copy = input_grid_b_copy.copy()
%store s3_input_grid_b_copy
Stored 's3_input_grid_b_copy' (DataFrame)